Chinese AI splits White House, Anthropic pays record copyright settlement

The launch of the free open-source model Kimi K3 by Chinese company Moonshot AI has caused discord among US President Donald Trump's AI advisors. Several former and current advisors publicly exchanged insults: David Sachs called Anthropic's models "lobotomized" and "politically correct," while Emil Michael from the Pentagon referred to the head of OpenAI's strategic futures division as "the village idiot supreme." The source of disagreement is the inability to reach a consensus on what to do about Kimi K3, which is claimed to rival paid models from OpenAI and Anthropic in intelligence. Each new intelligent free Chinese AI reduces Americans' incentive to pay for Western models, creating economic and political problems for the president. Meanwhile, a court approved Anthropic's record $1.5 billion settlement in a copyright infringement lawsuit: plaintiffs argued that the company used pirated works to train its Claude model. This is the largest known copyright payout in history, although many authors and creators do not consider it a victory. Additionally, the Trump administration is considering banning Chinese AI models, while China itself is contemplating tightening export controls on models and chips.